Not for a small business, and a proposal that suggests moving your SEO budget into GEO deserves hard questions. Every genuinely new GEO task, from prompt tracking to earned mentions, assumes you already have pages and listings worth citing. If those are missing, fix them first. The GEO layer can wait.

Can I check for free whether ChatGPT mentions my business?

Roughly, yes. Write down ten questions a customer would ask, such as "best roofer in [your town] for storm damage", and ask each one in ChatGPT, Perplexity and Google a few times, logged out. Note whether you are named and which websites are cited. Those cited sites tell you where a mention would matter most, which is useful before you talk to any agency.

Should I block AI crawlers on my website?

Block training if you like, but think twice before blocking search. OpenAI's training crawler, GPTBot, is separate from the one behind ChatGPT search, and OpenAI's crawler page says each setting is independent. So you can disallow GPTBot in robots.txt, refuse training, and still stay visible in answers.

GEO vs SEO: What the AI Search Line in Your Proposal Sells

GEO vs SEO comes down to where your business shows up. SEO (search engine optimization) works to get your pages ranked in Google’s list of results. GEO, short for generative engine optimization, works to get your business named or cited inside the answers that AI tools write, such as ChatGPT, Perplexity and Google’s AI Overviews. It has nothing to do with “geo” as in geography or local targeting.

The practical answer for a small business owner: most of what agencies now sell as GEO is ordinary SEO work with a new label, and a small part of it is genuinely new. Below I separate the two, show what published GEO prices look like and who publishes them, and give you five questions to ask before you pay extra.

GEO vs SEO: what each one actually means

SEO works to rank your pages in Google’s list of links, while GEO (generative engine optimization) works to get your business named or cited inside the answers that ChatGPT, Perplexity and Google’s AI Overviews write. The term comes from a 2023 research paper, GEO: Generative Engine Optimization, later accepted to the KDD 2024 conference. It described GEO as a way to improve “content visibility in generative engine responses” and reported that its methods could lift visibility “by up to 40%” in the researchers’ own tests. Agencies adopted the name quickly. The test conditions did not travel with it.

Here is the difference in plain terms:

	SEO	GEO
Where you appear	A ranked list of links	Inside a written answer, as a mention or a cited source
What counts as a win	A higher position, then a click	Being named, recommended or linked in the answer
How stable it is	Positions move slowly	The same question can get a different answer an hour later
How you measure it	Search Console, analytics, rank tracking	Running a fixed set of questions again and again and counting mentions

One warning before you read your own proposal. “Geo” in an agency document can also mean geographic work: city pages, “geo-targeted” landing pages, local listings. That is local SEO. If the proposal says “geo pages”, ask which kind of geo it means, because the two are priced and judged very differently.

What Google says about AI Overviews and SEO

Google says AI Overviews need no extra work beyond ordinary SEO, so for Google’s own AI features, GEO is SEO. Its guide to AI features and your website states: “There are no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary.” It adds that “the best practices for SEO remain relevant” and that you “don’t need to create new machine readable files, AI text files, or markup” or add any special schema.

Any line item promising special AI Overviews optimization is selling the same work under a different heading.

The same guide says traffic from AI features is included in your overall search traffic in Search Console, in the Performance report under the “Web” search type. It is counted with everything else, not reported on its own. Keep that in mind when an agency reports “AI Overviews traffic”: ask where the number came from.

ChatGPT and Perplexity are separate systems, and there is one real technical point here. OpenAI’s crawler documentation says sites that block its search crawler, OAI-SearchBot, “will not be shown in ChatGPT search answers.” Checking that your robots.txt and your hosting firewall do not block these crawlers is worth doing. It is a one-off check that takes minutes, not a monthly service.

Which GEO tasks are genuinely new, and which are SEO with a new label

Only four GEO tasks are genuinely new: tracking AI mentions on a written list of questions, earning brand mentions on sites AI answers draw from, keeping your business facts identical everywhere, and a one-off crawler check. Content, schema, reviews, Business Profile work and links are real and useful, but they are SEO, and Google’s AI features guide confirms that “the best practices for SEO remain relevant” for its AI answers. Price them once, in the SEO scope. This is the sort I do first on every AI line item in a proposal.

Genuinely new	SEO with a new label
Tracking whether AI answers name or cite you, on a written list of questions	Blog posts, FAQ pages, “answer-first” content
Earning mentions of your brand on sites AI answers draw from (industry lists, publications, comparison pages), done openly	Schema and structured data
Keeping your name, category and business facts identical across your site, Google Business Profile, directories and social profiles	Reviews and review responses
A one-off check that AI crawlers are not blocked	Google Business Profile, citations, link building, page speed, author pages

The risk is not that the right-hand work is bad. The risk is paying for it twice: once inside the monthly SEO retainer and again inside an “AI visibility package” that lists the same tasks. Put the two task lists side by side and mark every duplicate. If the AI section contains nothing from the left column, it is SEO, and it should be judged as SEO against the standard list of SEO services.

SEO vs AEO: the same shelf with another label

AEO (answer engine optimization) is another name for GEO, not a separate discipline: both mean shaping your content so AI tools and answer boxes quote it. Some agencies say AEO, some say GEO, some say LLMO (large language model optimization) or “AI visibility”. The label tells you about the agency’s marketing. It tells you nothing about the tasks underneath.

So do not compare the names. Compare the task lists. An “AEO package” with FAQ pages and schema is the right-hand column above. An “AEO package” with a prompt list, a baseline and outreach for mentions contains something new.

Google’s John Mueller put the sales pattern bluntly in August 2025: “The higher the urgency, and the stronger the push of new acronyms, the more likely they’re just making spam and scamming” (Search Engine Roundtable). A new acronym is not a scam by itself. A new acronym plus a deadline to sign deserves a slow read.

Those same AI answers are why so many owners now ask is SEO worth it for a small business at all. For searches where people pick a business to call, the answer has changed less than the new acronyms suggest.

How agencies sell GEO in a proposal

Agencies sell GEO in one of four shapes, and the riskiest is an add-on priced on top of a retainer you already pay:

A line inside the SEO retainer. “AI search optimization” listed with the other services, no separate fee. Usually harmless. Check it is counted like any other line.
An add-on priced on top. A separate “AI visibility” fee added to an existing retainer. This is where double billing happens most.
A whole GEO proposal. Every deliverable is AI work. Here the question is whether any of it can be measured at all.
An “AI readiness” audit or score. A one-off fee for a score from the agency’s own tool, with no method explained.

Three items in these proposals deserve extra care. A promise to “rank you in ChatGPT” or “guarantee AI citations” is the same trick as guaranteed rankings, and an even weaker promise, because AI answers change from one run to the next.

An llms.txt file sold as a headline deliverable is filler: it takes minutes to write, Google says no AI text files are needed, and Mueller wrote in June 2025 that “no AI system currently uses llms.txt” (Search Engine Roundtable).

Any plan to “seed” mentions on Reddit, Quora or review sites should say whose accounts post them. Mentions from made-up accounts, or paid mentions that are not labelled as paid, put your name at risk, not the agency’s.

How much does AI SEO cost?

Published GEO prices for small businesses run $1,500 to $5,000 a month, with projects at $5,000 to $50,000 and AI visibility tracking software at $50 to $1,000 or more a month, according to WebFX’s GEO cost guide. Look at who publishes them, though: WebFX is an agency that sells GEO.

Nutshell’s GEO pricing article prints the same ranges, tier for tier. That looks like a second opinion.

It is not. WebFX bought Nutshell in 2022. Two of the most-quoted GEO price guides are one seller’s numbers, published twice.

First Page Sage’s GEO cost breakdown is more revealing about what the money buys. It lists monthly tiers of $2,000 to $3,000, $4,000 to $7,000 and $8,000 to $12,000, drawn from “the small number of companies formally offering GEO services today”. Its cheapest tier “purchases low-cost placements on ranking websites”, and the middle tier publishes “superlative list articles”. In other words, part of what is sold as GEO is paid placement and self-published “best of” lists. Neither is wrong if it is disclosed. Both should be named as what they are.

I know of no independent survey of GEO prices yet. So I price an AI add-on the same way as any other line: count the genuinely new tasks, divide the fee by an hourly rate using the two-minute hours check, and see whether the hours match the work. If the add-on holds only relabeled SEO, its fair price is zero, because you are already paying for that work.

What a fair GEO line item should list and report

A fair GEO line item lists an agreed prompt list, the engines named, several runs per prompt, a month-one baseline with raw answers, and named mention targets, so you can check it without trusting the agency’s word. In detail:

A written prompt list, agreed before work starts. For a small business, a few dozen questions real customers would ask, fixed so the agency cannot pick winners afterward.
The engines named. ChatGPT, Perplexity, Google AI Overviews, Gemini or Copilot: which ones, and why those.
Several runs per prompt. AI answers are unstable. SparkToro’s 2026 study, in which 600 volunteers ran 12 prompts 2,961 times, found “a <1 in 100 chance” that ChatGPT or Google’s AI gives the same list of brands twice. Its conclusion: a visibility percentage across many runs is a reasonable metric, and “any tool that gives a ‘ranking position in AI’ is full of baloney.”
A month-one baseline and the raw answers shared with you, not a few screenshots.
Named targets for mentions, by type of site, with a quantity per quarter and paid placements labelled.
A cadence and a finish line for one-off jobs such as the entity cleanup and the crawler check.

Then each monthly AI-visibility report should show:

How often you were mentioned, per engine, as a share of all runs of the agreed prompts.
Which of your pages were cited, and for which prompts.
Which competitors appeared in the same answers.
The change against the month-one baseline.
The work done that month to earn those mentions.

What cannot be measured: “improve your AI presence”, “optimize for LLMs”, “make your brand AI-friendly”, or a score from a tool whose method nobody will explain. If a line reads like that, it cannot fail, so it cannot be checked.

Five questions to ask before paying extra for GEO

Before paying extra for GEO, ask which AI tasks are not already in the SEO scope, how tracking works, where mentions come from, whether the add-on can be dropped alone, and whether you get raw answers. Send them by email, so the answers are in writing:

“Which tasks in the AI section are not already in the SEO section?” Anything that appears in both should be priced once.
“Which prompts will you track, in which engines, how many times each, and what is my baseline?”
“Where will mentions of my business come from, under whose accounts, and which ones are paid?”
“Can I drop the AI add-on on its own, with 30 days’ notice, and keep the SEO work?”
“Will I get the raw answers every month, or only a summary score?”

A good agency can answer all five in writing, without another sales call. Vague answers to questions 2 and 3 usually mean the tracking has not been designed yet and you would be paying for it to be invented. Both matter for a reason: SparkToro found less than a 1 in 100 chance that ChatGPT or Google’s AI repeats the same list of brands, so tracking needs many runs, and First Page Sage’s cheapest GEO tier “purchases low-cost placements on ranking websites”, so you need to know which mentions are paid.

I audit GEO and AI-search proposals the same way as SEO ones, with five extra checks for this kind of line item. If your proposal has an AI section and you cannot tell what is new in it, send it to me before you sign.